Vous êtes sur la page 1sur 13
ARTICLE IN PRESS Global Environmental Change 16 (2006) 182–194 <a href=w w w . e l s e v i e r . c o m / l o c a t e / g l o e n v c h a Climate agreements based on responsibility for global warming: Periodic updating, policy choices, and regional costs Nathan Rive , Asbjørn Torvanger, Jan S. Fuglestvedt Center for International Climate and Environmental Research—Oslo (CICERO), P.O. Box 1129, Blindern, N-0318 Oslo, Norway Received 15 April 2005; received in revised form 16 December 2005; accepted 10 January 2006 Abstract It has been suggested that calculations of historical responsibility for global warming should be used to distribute mitigation requirements in future climate agreements. For a medium-term mitigation scenario, we calculate regional mitigation costs resulting from global allocation schemes based on the Brazilian Proposal that solely incorporate historical responsibility as a burden sharing criterion. We find that they are likely to violate ability-to-pay principles. In spite of less stringent abatement requirements, developing country regions experience cost burdens (as a percentage of GDP) in the same range as those of developed countries. We also assess the policy options available for calculating historical responsibility. The periodic updating of responsibility calculations over time, concerns over the robustness and availability of emissions data, and the question of whether past emissions were knowingly harmful, may lead to policy choices that increase the relative historical responsibility attributed to developing countries. This, in turn, would increase their mitigation cost burden. r 2006 Elsevier Ltd. All rights reserved. Keywords: Regional historical emissions; Contribution to global climate change; Brazilian Proposal; Dynamic CGE modelling; Regional mitigation costs 1. Introduction In climate policy, there exist numerous criteria which may be used to distribute emissions reduction requirements to participating countries. The Brazilian Proposal ( UNFCCC, 1997a ), presented prior to COP 3 in Kyoto, recommends allocating future emissions reduction require- ments among Annex I participants based on their historical contribution to climate change. The scheme could con- tribute favorably to the UNFCCC principle of ‘‘common but differentiated responsibilities’’ between Annex I and non-Annex I Parties ( UNFCCC, 1992 ) and thus received significant attention in the recent literature ( Rosa and Ribeiro, 2001 ; den Elzen and Schaeffer, 2002 ; den Elzen et al., 2005a, b ; Ho¨hne and Blok, 2005 ; Trudinger and Enting, 2005 ). The larger part of this attention has gone towards the key scientific and policy-related options in the calculation Corresponding author. Tel.: +47 22 85 87 50; fax: +47 22 85 87 51. E-mail address: nathan.rive@cicero.uio.no (N. Rive). 0959-3780/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.gloenvcha.2006.01.002 of regional historical responsibility (HR). The choice of climate change indicator, the emissions time period, the year in which climate change is evaluated, the greenhouse gas (GHG) inclusion, land-use and land-use change (LULUCF) emissions inclusion, attribution method, choice of climate model and the representation of climate processes—all influence the responsibility attributed to each region. The calculations can be particularly sensitive to a number of these options. If implemented in a burden sharing scheme, these policy choices would impact the cost burdens faced by each country, and alternative policy choices would generate relative ‘‘winners’’ and ‘‘losers’’. If an HR-based criterion is to be implemented in a future burden sharing schemes, some level of negotiation would be expected in order to agree upon these scientific and policy-related choices. The objective of this paper is to assess the relevant factors that may influence the choice of policy-related options in the calculations of HR, and highlight the implications for burden sharing in a future climate agreement. The factors we consider are fairness principles, operationalization issues related to data, and the periodic " id="pdf-obj-0-2" src="pdf-obj-0-2.jpg">

ARTICLE IN PRESS

ARTICLE IN PRESS Global Environmental Change 16 (2006) 182–194 <a href=w w w . e l s e v i e r . c o m / l o c a t e / g l o e n v c h a Climate agreements based on responsibility for global warming: Periodic updating, policy choices, and regional costs Nathan Rive , Asbjørn Torvanger, Jan S. Fuglestvedt Center for International Climate and Environmental Research—Oslo (CICERO), P.O. Box 1129, Blindern, N-0318 Oslo, Norway Received 15 April 2005; received in revised form 16 December 2005; accepted 10 January 2006 Abstract It has been suggested that calculations of historical responsibility for global warming should be used to distribute mitigation requirements in future climate agreements. For a medium-term mitigation scenario, we calculate regional mitigation costs resulting from global allocation schemes based on the Brazilian Proposal that solely incorporate historical responsibility as a burden sharing criterion. We find that they are likely to violate ability-to-pay principles. In spite of less stringent abatement requirements, developing country regions experience cost burdens (as a percentage of GDP) in the same range as those of developed countries. We also assess the policy options available for calculating historical responsibility. The periodic updating of responsibility calculations over time, concerns over the robustness and availability of emissions data, and the question of whether past emissions were knowingly harmful, may lead to policy choices that increase the relative historical responsibility attributed to developing countries. This, in turn, would increase their mitigation cost burden. r 2006 Elsevier Ltd. All rights reserved. Keywords: Regional historical emissions; Contribution to global climate change; Brazilian Proposal; Dynamic CGE modelling; Regional mitigation costs 1. Introduction In climate policy, there exist numerous criteria which may be used to distribute emissions reduction requirements to participating countries. The Brazilian Proposal ( UNFCCC, 1997a ), presented prior to COP 3 in Kyoto, recommends allocating future emissions reduction require- ments among Annex I participants based on their historical contribution to climate change. The scheme could con- tribute favorably to the UNFCCC principle of ‘‘common but differentiated responsibilities’’ between Annex I and non-Annex I Parties ( UNFCCC, 1992 ) and thus received significant attention in the recent literature ( Rosa and Ribeiro, 2001 ; den Elzen and Schaeffer, 2002 ; den Elzen et al., 2005a, b ; Ho¨hne and Blok, 2005 ; Trudinger and Enting, 2005 ). The larger part of this attention has gone towards the key scientific and policy-related options in the calculation Corresponding author. Tel.: +47 22 85 87 50; fax: +47 22 85 87 51. E-mail address: nathan.rive@cicero.uio.no (N. Rive). 0959-3780/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.gloenvcha.2006.01.002 of regional historical responsibility (HR). The choice of climate change indicator, the emissions time period, the year in which climate change is evaluated, the greenhouse gas (GHG) inclusion, land-use and land-use change (LULUCF) emissions inclusion, attribution method, choice of climate model and the representation of climate processes—all influence the responsibility attributed to each region. The calculations can be particularly sensitive to a number of these options. If implemented in a burden sharing scheme, these policy choices would impact the cost burdens faced by each country, and alternative policy choices would generate relative ‘‘winners’’ and ‘‘losers’’. If an HR-based criterion is to be implemented in a future burden sharing schemes, some level of negotiation would be expected in order to agree upon these scientific and policy-related choices. The objective of this paper is to assess the relevant factors that may influence the choice of policy-related options in the calculations of HR, and highlight the implications for burden sharing in a future climate agreement. The factors we consider are fairness principles, operationalization issues related to data, and the periodic " id="pdf-obj-0-6" src="pdf-obj-0-6.jpg">

Global Environmental Change 16 (2006) 182–194

ARTICLE IN PRESS Global Environmental Change 16 (2006) 182–194 <a href=w w w . e l s e v i e r . c o m / l o c a t e / g l o e n v c h a Climate agreements based on responsibility for global warming: Periodic updating, policy choices, and regional costs Nathan Rive , Asbjørn Torvanger, Jan S. Fuglestvedt Center for International Climate and Environmental Research—Oslo (CICERO), P.O. Box 1129, Blindern, N-0318 Oslo, Norway Received 15 April 2005; received in revised form 16 December 2005; accepted 10 January 2006 Abstract It has been suggested that calculations of historical responsibility for global warming should be used to distribute mitigation requirements in future climate agreements. For a medium-term mitigation scenario, we calculate regional mitigation costs resulting from global allocation schemes based on the Brazilian Proposal that solely incorporate historical responsibility as a burden sharing criterion. We find that they are likely to violate ability-to-pay principles. In spite of less stringent abatement requirements, developing country regions experience cost burdens (as a percentage of GDP) in the same range as those of developed countries. We also assess the policy options available for calculating historical responsibility. The periodic updating of responsibility calculations over time, concerns over the robustness and availability of emissions data, and the question of whether past emissions were knowingly harmful, may lead to policy choices that increase the relative historical responsibility attributed to developing countries. This, in turn, would increase their mitigation cost burden. r 2006 Elsevier Ltd. All rights reserved. Keywords: Regional historical emissions; Contribution to global climate change; Brazilian Proposal; Dynamic CGE modelling; Regional mitigation costs 1. Introduction In climate policy, there exist numerous criteria which may be used to distribute emissions reduction requirements to participating countries. The Brazilian Proposal ( UNFCCC, 1997a ), presented prior to COP 3 in Kyoto, recommends allocating future emissions reduction require- ments among Annex I participants based on their historical contribution to climate change. The scheme could con- tribute favorably to the UNFCCC principle of ‘‘common but differentiated responsibilities’’ between Annex I and non-Annex I Parties ( UNFCCC, 1992 ) and thus received significant attention in the recent literature ( Rosa and Ribeiro, 2001 ; den Elzen and Schaeffer, 2002 ; den Elzen et al., 2005a, b ; Ho¨hne and Blok, 2005 ; Trudinger and Enting, 2005 ). The larger part of this attention has gone towards the key scientific and policy-related options in the calculation Corresponding author. Tel.: +47 22 85 87 50; fax: +47 22 85 87 51. E-mail address: nathan.rive@cicero.uio.no (N. Rive). 0959-3780/$ - see front matter r 2006 Elsevier Ltd. All rights reserved. doi:10.1016/j.gloenvcha.2006.01.002 of regional historical responsibility (HR). The choice of climate change indicator, the emissions time period, the year in which climate change is evaluated, the greenhouse gas (GHG) inclusion, land-use and land-use change (LULUCF) emissions inclusion, attribution method, choice of climate model and the representation of climate processes—all influence the responsibility attributed to each region. The calculations can be particularly sensitive to a number of these options. If implemented in a burden sharing scheme, these policy choices would impact the cost burdens faced by each country, and alternative policy choices would generate relative ‘‘winners’’ and ‘‘losers’’. If an HR-based criterion is to be implemented in a future burden sharing schemes, some level of negotiation would be expected in order to agree upon these scientific and policy-related choices. The objective of this paper is to assess the relevant factors that may influence the choice of policy-related options in the calculations of HR, and highlight the implications for burden sharing in a future climate agreement. The factors we consider are fairness principles, operationalization issues related to data, and the periodic " id="pdf-obj-0-12" src="pdf-obj-0-12.jpg">

Climate agreements based on responsibility for global warming:

Periodic updating, policy choices, and regional costs

Nathan Rive , Asbjørn Torvanger, Jan S. Fuglestvedt

Center for International Climate and Environmental Research—Oslo (CICERO), P.O. Box 1129, Blindern, N-0318 Oslo, Norway

Received 15 April 2005; received in revised form 16 December 2005; accepted 10 January 2006

Abstract

It has been suggested that calculations of historical responsibility for global warming should be used to distribute mitigation requirements in future climate agreements. For a medium-term mitigation scenario, we calculate regional mitigation costs resulting from global allocation schemes based on the Brazilian Proposal that solely incorporate historical responsibility as a burden sharing criterion. We find that they are likely to violate ability-to-pay principles. In spite of less stringent abatement requirements, developing country regions experience cost burdens (as a percentage of GDP) in the same range as those of developed countries. We also assess the policy options available for calculating historical responsibility. The periodic updating of responsibility calculations over time, concerns over the robustness and availability of emissions data, and the question of whether past emissions were knowingly harmful, may lead to policy choices that increase the relative historical responsibility attributed to developing countries. This, in turn, would increase their mitigation cost burden. r 2006 Elsevier Ltd. All rights reserved.

Keywords: Regional historical emissions; Contribution to global climate change; Brazilian Proposal; Dynamic CGE modelling; Regional mitigation costs

1. Introduction

In climate policy, there exist numerous criteria which may be used to distribute emissions reduction requirements to participating countries. The Brazilian Proposal (UNFCCC, 1997a), presented prior to COP 3 in Kyoto, recommends allocating future emissions reduction require- ments among Annex I participants based on their historical contribution to climate change. The scheme could con- tribute favorably to the UNFCCC principle of ‘‘common but differentiated responsibilities’’ between Annex I and non-Annex I Parties (UNFCCC, 1992) and thus received significant attention in the recent literature (Rosa and Ribeiro, 2001; den Elzen and Schaeffer, 2002; den Elzen et al., 2005a, b; Ho¨hne and Blok, 2005; Trudinger and Enting, 2005). The larger part of this attention has gone towards the key scientific and policy-related options in the calculation

Corresponding author. Tel.: +47 22 85 87 50; fax: +47 22 85 87 51. E-mail address: nathan.rive@cicero.uio.no (N. Rive).

0959-3780/$ - see front matter r 2006 Elsevier Ltd. All rights reserved.

doi:10.1016/j.gloenvcha.2006.01.002

of regional historical responsibility (HR). The choice of climate change indicator, the emissions time period, the year in which climate change is evaluated, the greenhouse gas (GHG) inclusion, land-use and land-use change (LULUCF) emissions inclusion, attribution method, choice of climate model and the representation of climate processes—all influence the responsibility attributed to each region. The calculations can be particularly sensitive to a number of these options. If implemented in a burden sharing scheme, these policy choices would impact the cost burdens faced by each country, and alternative policy choices would generate relative ‘‘winners’’ and ‘‘losers’’. If an HR-based criterion is to be implemented in a future burden sharing schemes, some level of negotiation would be expected in order to agree upon these scientific and policy-related choices. The objective of this paper is to assess the relevant factors that may influence the choice of policy-related options in the calculations of HR, and highlight the implications for burden sharing in a future climate agreement. The factors we consider are fairness principles, operationalization issues related to data, and the periodic

ARTICLE IN PRESS

N. Rive et al. / Global Environmental Change 16 (2006) 182–194

183

updating of climate agreements and associated HR-based allocations. The most notable contributions of this paper to the literature are the use of (i) a global mitigation scenario that satisfies a long-term climate goal to 2100 and beyond, (ii) dynamically updated HR calculations to allocate emission reductions over time, and (iii) a computable general equilibrium (CGE) model to assess the regional mitigation cost impacts of alternative policy-related options. In the next section, we review the policy and scientific choices related to HR calculations. In Section 3, we detail factors that may influence these policy choices. In Section 4, we allocate emission reductions across eight regions on the basis of alternative HR calculations for a medium-term mitigation scenario. In Section 5, we discuss the results of these allocations, and compare the cost burdens faced by individual regions. In Section 6, we undertake a sensitivity analysis, and conclude with a summary of the main findings and implications for future burden sharing schemes in Section 7.

2. Brazilian Proposal policy-related options

Scientific and policy-related choices and their impact on the calculation of HR have been the main focus of the literature on the Brazilian Proposal. We briefly review these choices. HR is calculated for each country or region using an assumed period for attributed emissions—delineated by a chosen start and end years. Only emissions that occurred between these years will be used to assess each country’s responsibility. Emissions prior to or after this interval will not be attributed, but will be a part of the background emissions that contribute to the total climatic change. The evaluation year is the year in which the attributed emissions’ effect on the chosen climate indicator is measured. The evaluation year must obviously not precede the end of the attribution period. Ideally, we would seek to attribute responsibility for monetary damages arising climate change. However, given the uncertainty of future impacts, and the difficulty in monetizing these impacts, we use climate indicators as a proxy. In principle, any indicator on the climate change causal chain (i.e. cumulative emissions, radiative forcing, temperature, and sea level rise) can be used as an indicator for attributing HR. Global mean temperature was sug- gested in the original Brazilian Proposal (UNFCCC, 1997a), and is most typically considered in the HR literature. GHGs/forcing agents that are included in calculation of HR will affect the results, as the mix of GHGs emitted varies significantly across regions. The range of options includes using (i) fossil fuel CO 2 emissions only, (ii) fossil fuel and LULUCF CO 2 emissions only, (iii) man-made CO 2 , N 2 O, CH 4 emissions only, (iv) man-made Kyoto gas emissions, or (v) a wider range of man-made gases

including the Kyoto gases and precursors such as O 3 and SO 2 . The attribution method defines the methodology by which regional HR is calculated for the chosen start and end years, evaluation year, and climate indicator. There are several mathematical methods for attribution of responsi- bilities, which include the marginal, residual, and time- sliced methods. An in-depth discussion of these is made by Trudinger and Enting (2005). The impact of each of these policy-related options on the resulting HR calculation is varied (see den Elzen et al., 2005a). The choice of start and end years will decide whether the large backlog of developed country emissions will be included, and have a significant impact on their attributed HR. Extending the evaluation year far into the future will reduce the responsibility attributed to develop- ing countries, owing to their higher proportions of short- lived gases. Widening the types of emissions included in the calculation will also increase their attribution by taking into account LULUCF emissions. The choice of climate indicator has some impact, but it is less pronounced than the other choices, and the impact is dependent on the gas inclusion, and regional gas mix. Finally, the choice of attribution method is seen to have only a modest impact on the HR calculation.

3. Factors influencing policy-related options

It is impossible to know what form an HR-based burden sharing scheme could take in a future climate agreement, if it were employed at all. As discussed above, alternative policy-related options for HR calculation may generate relative advantages and disadvantages for regions partici- pating in a climate regime. As such, a level of negotiation would be required to agree upon a final set of policy options, which is likely to be influenced by a number of factors. Firstly, the chosen policy-related options will depend on how the HR calculations are to be implemented in the future regime. This, in turn, depends on the regime’s broader design. This includes the goals of the agreement, the modes of participation, the participating countries, and burden sharing (Philibert et al., 2003; Schmalensee, 1996; Philibert and Pershing, 2001; Torvanger and Ringius, 2002; Aldy et al., 2003; Torvanger et al., 2004a). Secondly, the policy-related choices may simply be adjusted to obtain desired HR and burden sharing outcomes. For example, the choice of a specific evaluation year may be largely arbitrary, and simply adjusted so that the resulting HR calculation is acceptable to the negotiating parties. Given that these two factors are specific to the characteristics of the future climate regime, we focus instead on a third set of ‘‘input-side’’ concerns. They are termed input-side as they define the bounds in which the alternative HR policy options are considered and nego- tiated. In doing so, these factors will undoubtedly influence the design of a future HR scheme. We identify three such

ARTICLE IN PRESS

  • 184 N. Rive et al. / Global Environmental Change 16 (2006) 182–194

input-side factors: fairness principles, operationalization issues, and the periodic updating of climate agreements and associated HR allocations. They are described as follows:

All burden sharing schemes embody a fairness principle to justify its use. For example, an HR scheme embodies the polluter pays principle—by allocating mitigation require- ments to those responsible for warming. An alternative scheme, a per capita convergence scheme (Agarwal and Narain, 1991; see also Contraction and Convergence in Meyer, 2000), embodies the principle of egalitarian emission rights by progressing towards per capita emission allowances for all people over time. In the context of this study, however, we also consider the fairness principles embodied in the alternative policy-related options within an HR scheme. The choice of these options will be influenced by different interpretations and preferences for fairness. Operationalization issues may also be a contributing factor to what policy options are chosen. Specific policy option alternatives may be difficult to implement due to technical, data, or logistical constraints in the climate regime. As a result, these options could become ‘unavail- able’ when an HR-based scheme is negotiated in a future regime. Finally, the dynamics of a long-term climate strategy may also play a role, through periodic updating of HR calculations. Future mitigation agreements will likely be negotiated on a short-term basis, as with the Kyoto Protocol. This would allow for the incorporation of new participants into the climate regime, as well as updated knowledge of the climate system. A key element of HR- based climate agreements could be the periodic updating of HR calculations to account for recent GHG emissions in the period between agreements. In the remainder of this section, we discuss the influence of these three factors on each policy-related option for HR calculation. In this paper, our focus is on the impact of choice of start year, end year, evaluation year, and climate indicator.

3.1. Start year

The start year delineates the start of emissions attributed to each region. A start year of 1890 is typically chosen, to take into account the backlog of emissions that contributed to current warming. However, the choice of a given start year could be thought of as embodying fairness principles. Firstly, the choice of a start year reflects an identification of who was emitting the GHGs, and thus who should undertake mitigation action in the climate agreement. The geopolitical map has changed significantly and roughly six generations have passed since 1890 (the typical choice for start year). Countries such as Russia and Germany were not in their current form prior to 1990. More critically, however, one could easily question whether the current generation should be held responsible for actions of their ancestors. The organization of a society’s economic and energy structure is largely inherited, and it has been shown

to be difficult to break out of the fossil energy ‘‘lock-in’’ (Unruh, 2000). Thus, by choosing a start year far into the past, we are applying a view of intergenerational fairness wherein populations can be considered collectively culp- able across both space and time. Secondly, a chosen start year also reflects a judgment that climate change was known about in the past, and that action could have been taken to avoid it. However, consensus on the causes and dangers of climate change has not spanned back many decades—certainly not to 1890. The roots of climate change theory are in the late 19th century, when Arrhenius (1896) postulated that increases in CO 2 concentrations in the atmosphere could lead to a rise in global mean temperatures. Revelle and Suess (1957) showed that anthropogenic CO 2 was unlikely to be fully absorbed by the ocean, and would likely cause warming of the planet. In 1990, the Intergovernmental Panel on Climate Change (IPCC) First Assessment Report (IPCC, 1990) concluded that man-made activities would lead to a warming in the next century of 0.3 1C per decade, yet mitigation obligations were only introduced by the Kyoto Protocol (UNFCCC, 1997b). In light of this timeline, it is difficult to decide a start year that would appropriately reflect the start of a nation’s fault in contributing towards warming. Beyond these fairness issues, the choice of start year is also affected by operationalization issues. HR calculations are heavily data-dependent. Regional and country-level responsibility calculations require region- and country- specific historical emissions scenarios. This gives rise to two specific data-related concerns. The first of these is data availability. The resolution of the responsibility calculations (how country-specific they can be) is ultimately limited by the resolution featured in the emission data. The EDGAR-HYDE 1.4 database (Van Aardenne et al., 2001; adjusted to Olivier and Berdowski, 2001), which provides our historical emissions profiles, has a disaggregation of 13 world regions, 1 and we are able to aggregate the EDGAR-HYDE regions to match the eight regions in our economic model, DEEP. 2 However, given the opportunity to implement further regions in the DEEP model, we would have been limited to the 13 region disaggregation found in the EDGAR-HYDE database. Of course, the EDGAR-HYDE is not the only historical emissions database available, but is the most comprehen- sive in terms of gas and region inclusion. And similar regional resolution limitations feature in the alternative GHG databases: The CDIAC database—which includes only fossil fuel (Marland et al., 2003) and land-use (Houghton, 1999) CO 2 emissions—has comprehensive data for the period 1890 to present for only 10 regions. Country- specific data is only available for fossil fuel emissions, with

1 Canada, USA, Latin America, Africa, OECD Europe, Eastern Europe, Former Soviet Union, Middle East, India, China, Southeast Asia, Oceania, Japan. 2 See Section 4 for details on the DEEP model.

ARTICLE IN PRESS

N. Rive et al. / Global Environmental Change 16 (2006) 182–194

185

varying historical time horizons. The IVIG database (de Campos et al., 2005) has country-level data, but only for land-use change emissions. The key point, then, is that our choice of database for HR calculation may restrict the number of regions, the choice of start year, and the number of gases included in the attribution. The second data concern relates to robustness. It has been shown that the calculation of HR is sensitive to the choice of historical emissions data (den Elzen et al., 2005a). While we do not examine this particular sensitivity in this study, it highlights the importance of an agreed-upon database for historical emissions, and an agreed methodology for reporting annual emissions. The methodology for reporting is provided by IPCC guidelines (IPCC, 1996). However, the annual reporting of inventories is only mandatory for Annex I countries. Annex I emissions inventories according to the IPCC guidelines for the six Kyoto gases are available from 1990 onwards. Data for non-Annex I countries (i.e. developing countries) for this period is not complete, as countries have reported at different times. Thus, while it is possible in future that a method for calculating historical emissions could be agreed upon internationally, using only the current (incomplete) IPCC inventories for HR calcula- tion would limit the start year to 1990. In Section 4, we calculate the impact on HR of three alternative start years: 1890, the standard year considered in the literature; 1945, marking the end of World War II; and 1990, roughly marking the end of the fall of Communism, the Rio Summit, and the start of IPCC GHG inventories.

3.2. End year

Like the start year, the end year also serves as a delineator of attributed emissions. However, there are arguably fewer fairness values embodied in the choice of this year. If climate action is to be taken in the near future, it is difficult to argue against using the most recent end year possible. The data issues surrounding the choice start year do not occur with the choice of end year provided an end year is chosen for which data is available. However, there are issues related to the dynamics of future climate agreements. It is likely that future post- Kyoto climate agreements will be updated at regular intervals. Thus, if HR remains a burden sharing criterion, the question arises as to whether the HR calculation should remain static to the end year used in the initial climate agreement, or be updated to account for the development in emissions in the period between negotiations. In Section 4, we compare the results of HR-based abatement allocations using a static end year with those using a dynamically updated end year. For the static end year, we choose 2000. This is largely arbitrary, as any single year would suffice for this comparison. In the dynamic cases, we use 10-year updating (an assumption of the time gap between climate negotiations), and 5-year updating (the maximum resolution of our CGE model).

  • 3.3. Indicator

Related to fairness principles, the choice of indicator may be influenced by country preferences towards indica- tors that are closely connected to the impacts they are concerned with. For example, a country like the Maldives may be particularly concerned with sea level rise, and prefer its use as indicator for HR calculations. Land-locked Switzerland may have a preference for temperature rise, given its impact on tourism. However, given that each of these climate indicators are precursors to the climate impacts we are ultimately concerned with, and that the UNFCCC (1992) does not specify which climate para- meters are most important, it is non-obvious which choice of indicator would be preferred. Related to operationalization, it has been argued that the chosen indicator should be robust to calculation, well understood, and discount the importance of emissions from a long time ago (Rosa et al., 2004). Global mean temperature satisfies these criteria as an indicator, and it was also recommended by the original Brazilian Proposal (UNFCCC, 1997a). It is widely used in other HR studies (e.g. den Elzen et al., 2005a). An alternative proposal is the use of CO 2 -equivalent cumulative GHG emissions as an indicator, given its ease of calculation, and simplicity of communication to policymakers (den Elzen et al., 2005b; Ho¨hne and Blok, 2005). However, this indicator fails to discount the emissions from a long time ago and the differing lifetimes of gases. In Section 4, we compare the use of global mean temperature change (DT) and cumulative emissions weighted by 100-year global warming potential (GWP) (Shine et al., 1990) on HR calculations. See den Elzen et al. (2005a) for a wider treatment of alternative indicators.

  • 3.4. Evaluation year, time gap, and background emissions

scenario (for DT indicator only)

The choice of evaluation year reflects the concern of when the climate change impacts become important for identifying historical responsibility. A decision must be made whether a gap is to separate the end and evaluation dates, and what size that gap should be. This is similar to the choice of time horizon in the application of GWPs. In choosing the evaluation year, it could be argued that we are more certain of climate change occurring today. The background emissions scenario and climate change in the future is associated with larger uncertainties, thus we may not be willing to set an evaluation year far into the future. Fairness concerns could argue that social time preferences place more emphasis on the now. On the other hand, it could be argued that the evaluation year should be placed some time in the future, at which time the climate system will have had time to react, and the climate impacts from our emissions will be more cumulative and visible. In Section 4, we examine the impacts on HR calcula- tions from both near-term (2000) and long-term (2100)

ARTICLE IN PRESS

  • 186 N. Rive et al. / Global Environmental Change 16 (2006) 182–194

Table 1 Policy-related options for calculating historical responsibility

Start year

1890, 1945, 1990

End year

2000, 10-year dynamic, 5-year dynamic

Indicator

GWP-weighted cumulative emissions,

Evaluation year

global mean temperature 2000, 2010, 2050, 2100

Background scenario

Case ME, zero, constant

evaluation dates. In addition, we examine the sensitivity to alternative end-year-evaluation year gaps (10 years, 50 years) for a fixed end year of emissions (2000). With an evaluation year far into the future, we also examine the sensitivity of HR calculations to the back- ground emissions scenario. This is important, as we do not know what the future emissions trajectories will look like, and we do not know how this will affect our HR calculations. In Section 4, we analyze scenarios where the background emissions after the end year either (i) are held at a constant level equal to the end year emissions, (ii) drop to zero, or (iii) follow our assumed global mitigation pathway (Case ME, described below). A summary of the policy-related options assessed in this study is shown in Table 1. (The default options are in bold.) For all our HR calculations, the emissions included are CH 4 , N 2 O, and CO 2 from industrial, agricultural, and household sources.

4. Allocating emission reductions

We focus on the medium-term period from 2000 to 2045, during which period we allocate emission reductions across eight world regions: China, Former Soviet Union (FSU), Eastern Europe, India, JCANZ region (Japan, Canada, Australia, and New Zealand), Western Europe, USA, and the Rest of World (ROW).

4.1. Emissions past and future

In this paper, the global emission reduction requirements (i.e. deviation from the BAU) which we allocate across the eight regions for the period 2000–2045 are taken from an existing long-term global mitigation scenario (from 2000 to 2145): the Case ME (Medium climate sensitivity, Early climate action) scenario generated in Torvanger et al. (2004b). This scenario satisfies a long-term global warming target of 2.5 1C above pre-industrial levels in 2100 (and a limited rise afterwards) for a climate sensitivity of 3.0 1C (for 2 CO 2 concentrations). 3 This long-term scenario also provides the background emissions for responsibility calculations when the evaluation year is far into the future.

3 When calculating HR in our climate model, the climate sensitivity is set at 3.0 1C for consistency. However, it is seen that HR calculations are insensitive to the choice of climate sensitivity (den Elzen et al., 2005a).

This scenario is shown in Fig. 1 (medium-term perspective) and Fig. 2 (long-term perspective) below. We use both the CICERO simple climate model (SCM) (Fuglestvedt and Berntsen, 1999) and the Dynamic analysis of Economics and Environmental Policy (DEEP) CGE model (Kallbekken, 2004; used in Kallbekken and Wests- kog, 2005) to calculate the regional HR, allocate emissions reduction requirements, and calculate mitigation cost burdens. The DEEP model uses economic growth and technological improvement parameters from the IPCC (2001) Special Report on Emissions Scenarios (SRES) A1B scenario, and thus generates a business as usual scenario with economic and emissions growth equal to that of the

16 14 12 10 8 6 4 "Case ME" Scenario 2 DEEP BAU (A1B parameters) 0
16
14
12
10
8
6
4
"Case ME" Scenario
2
DEEP BAU (A1B parameters)
0
2000
2010
2020
2030
2040
CO 2 Emissions (GtC/year)

Year

Fig. 1. Annual CO 2 emissions in our global mitigation scenario (Case ME) and the DEEP BAU scenario

20 18 16 14 12 10 8 6 SRES A1B 4 "Case ME" Scenario SRES B2
20
18
16
14
12
10
8
6
SRES A1B
4
"Case ME" Scenario
SRES B2
2
0
2000
2020
2040
2060
2080
2100
CO 2 Emissions (GtC/year)

Year

Fig. 2. Annual CO 2 emissions in our global mitigation scenario (Case ME) and SRES A1B and B2 scenarios. Source: IPCC (2001a).

ARTICLE IN PRESS

N. Rive et al. / Global Environmental Change 16 (2006) 182–194

187

A1B scenario. (See the Annex for further information on the models.) The historical emissions data for each region is taken from the EDGAR-HYDE 1.4 database (Van Aardenne et al., 2001, adjusted to Olivier and Berdowski, 2001) for the period 1890–1995. We join the EDGAR-HYDE database emissions in 1995 to the DEEP emissions in 2000 through linear interpolation. It should be noted that the choice of our eight regions was largely determined by regional specifications in the EDGAR-HYDE database, and the finite computational capabilities of the DEEP model solver. The EDGAR- HYDE database features 13 regions, of which only Canada, USA, Japan, China, and India are unique countries. The DEEP model is only able to include eight aggregated regions when it is run for the period 2000–2045.

4.2. GHG mitigation

The DEEP model includes emissions and abatement for industrial, agricultural, and domestic CO 2 , N 2 O, and CH 4 . As such, they are assumed to be the sole source of mitigation in the abatement scenario, and the only emissions included in the HR calculation. Emissions from the remaining KP gases (SF 6 and the F-gases) are assumed to follow the A1B business as usual scenario. This is acceptable because they are excluded from the HR calculation, and we find in Section 5 that background emissions have no influence on the calculation. LULUCF emissions are similarly excluded. These exclusions are dictated by DEEP, which is unable to model the emissions or abatement of SF 6 , the F-gases, and emissions from LULUCF sources. It should be noted that by excluding LULUCF sources, our HR calculations are somewhat biased towards higher developed country HR (compared to if LULUCF sources were included). The exclusion of SF 6 and the F-gases has only a small impact (see den Elzen et al., 2005a for the impact of gas mix on HR calculations). Overall, we find the exclusions acceptable given that industrial, agricultural, and domestic sources make up the bulk of national GHG emissions, and the accounting of emissions from LULUCF sources is highly uncertain.

  • 4.3. Attribution methodology

When calculating regional HR with global mean temperature change (DT) as indicator, we employ the normalized marginal method as described by Trudinger and Enting (2005) and den Elzen et al. (2005a). In short, each region’s marginal contribution to warming is simulated in the climate model. By calculating this warming for each region in turn, we are able to calculate the share of total global warming resulting from each region’s emissions.

  • 4.4. Allocation methodologies

In this paper, the global mitigation burden (the required deviation from the BAU emissions level, shown in Fig. 1) is allocated among participating countries for each commit- ment period. Table 2 lists the alternative HR-based allocation schemes examined in this study. For clarity, we introduce the definition of the allocation year in HR-based allocation schemes: the year in which the GHG abatement (allocated by an HR calculation) takes place. The use of this term ensures we differentiate the allocation year from the start, end, and evaluation years. We employ an allocation-based methodology (as defined in Rose et al. (1998)) to apply our HR calculations. A participant will be allocated a share of the global mitigation requirement (in terms of tons of GHG) equal to its calculated HR. For example, for a given allocation year, if the USA is calculated to have historically contributed to 20% of warming, it will be assigned 20% of global mitigation burden (in terms of CO 2 equivalent emission reductions) in the allocation year. Such a simple methodology is the simplest interpretation of the Brazilian Proposal principles; we keep separate alternative fairness criteria such as ability-to-pay or grandfathering. It is certainly feasible to merge an HR-based burden sharing scheme with other fairness criteria, such as ability-to-pay thresholds (e.g. den Elzen et al., 2005b). However, we choose not to do so, in order to assess the effects of using a burden sharing scheme that solely uses HR criteria. With regards to participation, we assume all countries, both developed and developing, will be participating in

Table 2 Policy-related HR paramenters in alternative experimental runs

Run

Start Year

End Year

Evaluation Year

Indicator

Background

1

1890

10-year dynamic

2100

Temperature

Case ME

2

1990

10-year dynamic

2100

Temperature

Case ME

3

1945

10-year dynamic

2100

Temperature

Case ME

4

1890

2000

2100

Temperature

Case ME

5

1890

5-year dynamic

2100

Temperature

Case ME

6

1890

2000

2000

Temperature

Case ME

7

1890

2000

2010

Temperature

Case ME

8

1890

2000

2050

Temperature

Case ME

9

1890

2000

Cumulative Emissions

Case ME

10

1890

2000

2100

Temperature

Zero

11

1890

2000

2100

Temperature

Constant

ARTICLE IN PRESS

  • 188 N. Rive et al. / Global Environmental Change 16 (2006) 182–194

mitigation in our climate regime. The role of developing countries in future climate agreements remains under discussion; however, it is highly likely that participation in future regimes will be expanded beyond the Annex-I grouping with some form of obligations. We assume that all countries will undertake quantitative emission reduc- tions, differentiated only by the HR calculations. Further- more, we allow full emission permit trading for all participants. The trading system is based on a GWP 100 exchange rate between the three gases. In our experiment, HR will be calculated from the initial emissions permit endowment granted to each country under the climate regime, rather than the number of post-trading permits held. This ensures that the calculations account for all mitigation paid for by each country, rather than the geographical location of that mitigation (as it may be bought from other countries to minimize the abatement costs).

5. Results and discussion

This section discusses the results of our HR-based allocation scenarios listed in Table 2. Run 1 serves as the reference case for comparison with the alternatives, although it should not be seen as the preferred set of policy-related parameters. The results of the HR calculations for Run 1 are shown in Fig. 3 for the mitigation period 2020–2045. It is a step function because it assumes that the HR is only updated (with a new end year) every 10 years, to simulate the regular re-negotiation of climate commit- ments. Thus, an HR calculation made with an end year of 2019 is used for allocations from 2020 through to 2029. The HR is then re-calculated for the 2030s with an end year of 2029, and so forth. It is seen that for the first years of

allocation, the US and ROW respectively have the highest shares of HR. However, beyond 2030, the calculations begin to take into account the strong emissions growth among the developing countries. A key result is the significant increases shown by HR attributed to China, India, and ROW over the mitigation period. The ROW region, which is made up of most of Asia, Africa, and South America, sees its share of global HR rise from 20% in 2020 to 27% in 2040. Similarly, China overtakes Western Europe in terms of share of global HR with a rise from 13% in 2020 to 16% in 2040. The results from Fig. 3 are iteratively used to allocate emission reductions across the eight regions (Fig. 4), with the resulting emissions then used to calculate HR for the next periodic update. All reduction requirements are increasing over time, as a result of the growing global abatement requirement (see Fig. 1). However, it is seen that developing countries are allocated a significantly lower percentage reduction from their BAU compared to developed countries, owing to their smaller shares of responsibility for warming (Fig. 4).

5.1. Mitigation costs

We use the DEEP model to calculate the annual total abatement cost for each country, as a percentage of their annual GDP. The total abatement cost is calculated by multiplying the average abatement costs (the opportunity cost of reducing emissions) by the regional emission reductions. Fig. 5 shows the cost results from Run 1, with the bars denoting the range of burden associated with all the alternative HR runs (see Table 2). With the exception of Eastern Europe and the FSU, the mitigation burden across all regions ranges from roughly 0.02% (ROW in Run 10) to 0.13% (India in Run 6) of

30% ROW FSU USA JCANZ 25% China India W Europe E Europe 20% 15% 10% 5%
30%
ROW
FSU
USA
JCANZ
25%
China
India
W Europe
E Europe
20%
15%
10%
5%
0%
2020
2025
2030
2035
2040
2045
Share of Global Warming

Allocation Year

Fig. 3. Run 1 historical responsibility results for period 2020–2045, in terms of share of global contribution to warming.

25% W Europe FSU E Europe India USA ROW 20% JCANZ China 15% 10% 5% 0%
25%
W Europe
FSU
E Europe
India
USA
ROW
20%
JCANZ
China
15%
10%
5%
0%
2020
2025
2030
2035
2040
2045
Reduction from regional BAU

Allocation Year

Fig. 4. Run 1 allocated annual emission reductions (% reduction from regional BAU emissions).

ARTICLE IN PRESS

N. Rive et al. / Global Environmental Change 16 (2006) 182–194

189

0.40 FSU 0.35 E Europe 0.30 0.25 0.20 0.15 India USA W Europe China 0.10 JCANZ
0.40
FSU
0.35
E Europe
0.30
0.25
0.20
0.15
India
USA
W Europe
China
0.10
JCANZ
ROW
0.05
0.00
Total Annual Abatement Cost in 2045 (% of GDP)

Fig. 5. Total annual abatement cost in 2045 by country (% of GDP) for Run 1. Bars denote range of burden allocated to each region under all alternative Runs (see Table 2).

regional GDP. 4 What is remarkable is that the cost burden of developing countries under our HR-based allocation scheme is in a range similar to that of developed countries, even though their abatement requirements (in terms of percentage reduction from BAU emissions) are less stringent. This is driven by the relatively higher carbon intensity of GDP seen in developing countries (see Annex Fig. A1). A similar result is seen by Zhang (2000), who finds GNP loss in China under less restrictive mitigation scenarios to be comparable to GNP loss in developed regions under stringent mitigation requirements. It should be noted that measuring abatement costs in terms of percentage GDP may be an insufficient means of measuring burden. In particular, it may be a poor representation of the mitigation burden faced by develop- ing countries relative to their developed country counter- parts, as it fails to take into account equity issues. A mitigation burden of 0.5% of GDP will be more severe to developing countries with lower per capita GDP (and thus higher marginal utility of consumption) than richer developed countries. This issue could be ameliorated through the use of a GDP per capita weighting system on calculated mitigation costs. However, the simplicity of percentage GDP cost reporting is sufficient for our conclusions. In the previous section, it was pointed out that alternative policy-related parameters in attributing HR can benefit either developing or developed countries. Yet

4 Eastern Europe and FSU are anomalous because of their high carbon intensity of GDP from industrial GHGs (see Fig. A1 in the Annex), and their proportionally high HR from Soviet-era GHG emissions.

our results in Fig. 5 show that the mitigation costs remain remarkably similar for both developed and developing

regions, regardless of the alternative policy-related choices that are used to calculate HR. This suggests that pure HR-

based allocation schemes considered in this study (which

do not employ alternative fairness criteria) may violate

ability-to-pay (or other equity) principles. This result is

only strengthened when considering equity issues: the

calculated mitigation cost burden to developing countries

will become significantly greater if an equity weighting is

applied. Consequently, it is unlikely that an HR-based

burden sharing scheme would be incorporated into a

future climate agreement in the format considered in this

paper.

Next, we briefly highlight the impact of the alternative

policy-related options for calculating HR on the regional

mitigation cost burdens.

  • 5.2. Start year

Moving from the default start year (1890) to more recent years (1990) for attributing emissions, we find that the burden on developing countries increases (see Fig. 6). This is because more recent start years exclude the long backlog of developed country long-lived GHG emissions. This is in agreement with den Elzen et al. (2005a).

  • 5.3. End year

By dynamically updating the end year forward, we find that developing country burdens are increased, while developed country burdens fall (see Fig. 7). This is a result of the fast-growing emissions likely to be seen in developing country economies over the coming decades— featured particularly in the SRES A1B scenario employed in this study. Over time, these emissions contribute to an increasing share of future warming. As a consequence, developing country HR will rise, and developing countries will be continuously allocated a larger piece of a growing global emissions reduction burden (in terms of tons of GHG). This is not a surprising result, but this type of periodically updated calculation has not been undertaken in the literature before.

  • 5.4. Evaluation year

Extending the gap between the end year (in this case, 2000) and evaluation year is seen to favor developing countries (Fig. 8). These regions have a higher proportion of methane in their emissions mix, whose atmospheric lifetime is relatively short (12 years). As the gap between end year and evaluation year is extended beyond the lifetime of methane, there is a significant drop in the developing country burden. This is in agreement with den Elzen et al. (2005a).

ARTICLE IN PRESS

  • 190 N. Rive et al. / Global Environmental Change 16 (2006) 182–194

0.3 FSU Start Year: 1890 0.25 Start Year: 1945 Start Year: 1990 E Europe 0.2 0.15
0.3
FSU
Start Year: 1890
0.25
Start Year: 1945
Start Year: 1990
E Europe
0.2
0.15
China
India
USA
0.1
W Europe
JCANZ
ROW
0.05
0
Total Abatement Cost in 2045 (% of GDP)
0.4 FSU Evaluation Year: 2000 Evaluation Year: 2010 0.35 Evaluation Year: 2050 E Europe Evaluation Year:
0.4
FSU
Evaluation Year: 2000
Evaluation Year: 2010
0.35
Evaluation Year: 2050
E Europe
Evaluation Year: 2100
0.3
0.25
0.2
0.15
India
USA
W Europe
0.1
JCANZ
China
0.05
ROW
0
Total Abatement Cost in 2045 (% of GDP)

Fig. 6. Regional total annual abatement

cost

(%

of

GDP) under

Fig. 8. Regional total

annual abatement cost

(%

of

GDP) under

alternative start year parameters: 1890 (Run 1), (Run 2).

1945 (Run

3),

1990

alternative evaluation year parameters: 2000 (Run 6), 2010 (Run 7), 2050 (Run 8), 2100 (Run 4).

0.4 FSU End Year: 2000 0.35 End Year: 10-year updated E Europe End Year: 5-year updated
0.4
FSU
End Year: 2000
0.35
End Year: 10-year updated
E Europe
End Year: 5-year updated
0.3
0.25
0.2
0.15
USA
India
W Europe
0.1
China
JCANZ
ROW
0.05
0
Total Abatement Cost in 2045 (% of GDP)

Fig. 7. Regional total annual abatement cost (% of GDP) under alternative end year parameters: 2000 (Run 4), 10-year dynamic (Run 1), 5-year dynamic (Run 5).

5.5. Indicator

Using cumulative emissions as a climate indicator is found to reasonably replicate the results from comparable runs evaluating DT at either 2000 or 2100 (Fig. 9; end year is 2000). This is in agreement with results from den Elzen et al. (2005a, b).

0.4 FSU Indicator: Cumulative Ems Indicator: 2000 dT 0.35 Indicator: 2100 dT E Europe 0.3 0.25
0.4
FSU
Indicator: Cumulative Ems
Indicator: 2000 dT
0.35
Indicator: 2100 dT
E Europe
0.3
0.25
0.2
0.15
India
USA
W Europe
0.1
China
JCANZ
0.05
ROW
0
Total Abatement Cost in 2045 (% of GDP)

Fig. 9. Regional total annual abatement cost (% of GDP) under alternative climate indicators: GWP-weighted cumulative emissions (Run 9), 2000 global mean temperature change (Run 6), 2100 global mean temperature change (Run 4).

5.6. Background emissions

We find negligible sensitivity to the choice of a post-2000 background scenario (assuming a static end year of 2000), illustrated in Fig. 10. This means that HR calculations will not be affected by the uncertain background emissions trajectories in future. This suggests that, if desired, it is

ARTICLE IN PRESS

N. Rive et al. / Global Environmental Change 16 (2006) 182–194

191

0.4 Bkgrnd Ems: "Case ME" FSU Bkgrnd Ems: Zero 0.35 Bkgrnd Ems: Constant E Europe 0.3
0.4
Bkgrnd Ems: "Case ME"
FSU
Bkgrnd Ems: Zero
0.35
Bkgrnd Ems: Constant
E Europe
0.3
0.25
0.2
0.15
USA
W Europe
0.1
India
JCANZ
China
0.05
ROW
0
Total Abatement Cost in 2045 (% of GDP)

Fig. 10. Regional total annual abatement cost (% of GDP) under alternative post-2000 background emissions: Case ME scenario emissions (Run 4), zero emissions (Run 10), constant emissions (Run 11).

operationally feasible to employ an evaluation year far into the future without knowing what future background emissions levels will be.

6. Sensitivity analysis

In this section, we undertake a brief sensitivity analysis using alternative economic growth and technological improvement parameters in the DEEP model. We are particularly interested in the sensitivity of the dynamically updated HR calculations to our economic growth and technological improvement assumptions. For the above runs, the DEEP model was set up using parameters from the SRES A1B scenario. The A1B scenario describes a high-growth future, with rapid emissions growth in the developing world. As an alternative, we consider the impact of employing economic and technological para- meters from the SRES B2 scenario (IPCC, 2001), which describes a regionalized and environmentally concerned future. In contrast to the A1B scenario, the B2 shows only modest developing country emissions growth over the coming decades (see Fig. 2 for reference). In this sensitivity analysis, we re-run our Run 1 (see Table 2), but this time with B2 parameters in the DEEP model. Our mitigation scenario uses the same percentage global emissions reductions (from the DEEP B2 BAU) as is required in the original scenarios (Fig. 1). The dynamic HR calculations are shown in Fig. 11. We find that with B2 parameters, the HR attributed to developing countries still increases over the period 2020–2045, while the developed country HR falls. However, the trends are less pronounced than with A1B economic and technological assumptions. The difference is seen in the ROW region results in

30% ROW FSU USA JCANZ 25% China India W Europe E Europe 20% 15% 10% 5%
30%
ROW
FSU
USA
JCANZ
25%
China
India
W Europe
E Europe
20%
15%
10%
5%
0%
2020
2025
2030
2035
2040
2045
Share of Global Warming

Allocation Year

Fig. 11. Historical responsibility calculations for period 2000–2045 in terms of share of global contribution to temperature change, using Run 1 parameters. Solid line denotes results using SRES A1B growth parameters in DEEP, dotted line denotes SRES B2 parameters.

0.3 FSU Run 1: SRES A1B Params Run 1: SRES B2 Params 0.25 E Europe 0.2
0.3
FSU
Run 1: SRES A1B Params
Run 1: SRES B2 Params
0.25
E Europe
0.2
0.15
India
USA
0.1
China
W Europe
JCANZ
0.05
ROW
0
Total Abatement Cost in 2045 (% of GDP)

Fig. 12. Regional total annual abatement cost (% of GDP) under Run 1, using A1B and B2 growth parameters in DEEP.

particular, owing to the slow emissions growth in the SRES Africa and Latin America (ALM) region under the B2 scenario. With regards to mitigation costs (in terms of percentage GDP), we see in Fig. 12 that under the B2 case, the developing country burdens remain in the same range as those of the developed countries. While the cost burden to all countries is generally smaller with B2 parameters compared to A1B parameters (a result of the differing

ARTICLE IN PRESS

  • 192 N. Rive et al. / Global Environmental Change 16 (2006) 182–194

economic assumptions), our key observation holds: under a simplistic HR-based burden scheme such as this, develop- ing and developed country mitigation cost burdens are not sufficiently differentiated, thereby likely violating ability- to-pay (or other equity) principles.

7. Conclusion

Results from this study, as well as those in the literature, show that historical responsibility (HR) calculations have a strong sensitivity to the policy-related choices used. While it is not possible to predict the choice among these options in a future climate agreement, we have discussed the factors that may influence the choice of HR policy options. The influencing factors we have identified (fairness principles, operationalization, and data issues) are seen to be useful evaluation criteria. Using an integrated climate and economic model, we have assessed the possible implica- tions of these influencing factors on regional mitigation cost burdens arising from an HR-based allocation scheme. This paper makes a number of findings. Firstly, a pure HR-based allocation scheme, as employed in this study, may impose cost burdens on developing countries that, compared to developed country cost burdens, may be unacceptably high. While less stringent emission reductions (in terms of tons of GHG emissions) are allocated to developing countries than to developed countries, their cost burdens (as a percentage of GDP) are roughly in the same range. The developing country burden is further increased when taking into account equity issues. This could violate ability-to-pay (and other equity) principles. This result appears to be robust across the range of possible policy- related choices in calculating HR, and future economic growth and technological improvement assumptions. Secondly, the choice of start year is likely to be influenced by a number of factors. Concerns over emissions data availability and robustness may push for a more recent start year in place of the typical choice of 1890. Furthermore, while we may wish to include a long backlog of emissions in acknowledgement of their cumulative warming impact, a problem still exists with regard to whether historical emissions were knowingly causing climate change. This may also push for a more recent start year. On the whole, these influencing factors will likely lower the HR attributed to developed countries. Third, the choice of end year will likely be influenced by the design and dynamics of the climate agreements. If future climate regimes are negotiated periodically (as with the Kyoto Protocol), HR may also be updated periodically. This causes developing countries’ attributed HR to rise over time, given their fast-growing greenhouse gas emis- sions. This requires them to take an increasing share of a growing global abatement burden over time. Fourth, the evaluation year is likely to be chosen with a sufficient gap into the future (from the end year) to account for climate inertia. Operationally, we find this to be feasible, as we find no sensitivity in the HR calculations

to the future emissions scenario. Sensitivity to the chosen gap depends on whether it is beyond the horizon of methane’s lifetime in the atmosphere (assuming, of course, that methane is used in the calculation). A larger gap decreases the HR of developing countries, which have a larger share of short-lived gases in their emissions profiles. Finally, the use of a cumulative CO 2 -equivalent emis- sions indicator generates a similar HR calculation to that of global mean temperature change, with a marginal reduc- tion to developing countries’ responsibility. As pointed out by den Elzen et al. (2005a), cumulative CO 2 -equivalent emissions may be a preferable indicator to temperature change as it allows a more simple calculation, and is easier to explain to policymakers. However, it may be preferable to keep temperature change as an indicator, given its close association with the climate impacts we seek to avoid, understanding by the general public, and flexibility with regard to alternative policy-related choices in the calcula- tions (e.g. evaluation year). In future burden sharing schemes, negotiators will likely seek to avoid violating ability-to-pay (or other equity) principles when allocating emissions reductions among countries. As such, the HR-based scheme as used in this study is likely to be infeasible in political terms. A possible solution could be a hybrid burden sharing scheme incorporating at least two of alternative fairness criteria (Rose et al., 1998; Berk and den Elzen, 2001), or a participation threshold for those taking on mitigation commitments based on GDP per capita (den Elzen et al., 2005b). The exploration of more promising combinations of HR and other allocation schemes is an important issue for future studies.

Acknowledgments

This work has received funding from the Norwegian Research Council’s RAMBU program, through the project A Sustainable Climate Development: Long-term Targets, Global Agreements, and Economic Consequences. The authors would like to thank Terje Berntsen, Michel den Elzen, Steffen Kallbekken, and Brian O’Neill, and two anonymous reviewers for their valuable comments.

Appendix A. Model description

The SCM is described in Fuglestvedt and Berntsen (1999) used in, e.g., Fuglestvedt et al. (2000). The model incorporates a scheme for CO 2 from Joos et al. (1996) and an energy-balance climate/up-welling diffusion ocean model developed by Schlesinger et al. (1992). The CO 2 module uses an ocean mixed-layer pulse response function that characterizes the surface to deep ocean mixing in combination with a separate equation describing the air–sea exchange based on the HILDA model (Siegenthaler and Joos, 1992) to account for non-linearities in the carbon chemistry in the ocean.

ARTICLE IN PRESS

N. Rive et al. / Global Environmental Change 16 (2006) 182–194 193

1.4 FSU 1.2 1.0 China E Europe India 0.8 0.6 0.4 ROW USA 0.2 JCANZ W
1.4
FSU
1.2
1.0
China
E Europe
India
0.8
0.6
0.4
ROW
USA
0.2
JCANZ
W Europe
0.0
Carbon (CO 2 ) intensity of GDP
(tC per 1000 1997US$GDP)

Fig. A1. Non-LULUCF carbon intensity of GDP in 2000. Source: GTAP and GTAP/EPA databases (Dimaranan and McDougall, 2002; Lee, 2002).

The DEEP model is described in Kallbekken (2004) and used in, e.g., Kallbekken and Westskog (2005). The DEEP model is a multi-sector, multi-region, multi-gas CGE model with a growth benchmark. The model was run for ten 5-year periods, from 2000 to 2045. The structure of production and demand in the DEEP model has been adopted from the GTAP-EG model by Rutherford and Paltsev (2000), and the social accounting matrix (SAM) is version 5 of the Global Trade Analysis Project (GTAP) database (Dimaranan and McDougall, 2002) (see Fig. A1). We use economic growth rates and technological improve- ment parameters from SRES A1B scenario (IPCC, 2001) to generate a ‘‘DEEP A1B’’ BAU scenario. There are no explicit emissions abatement curves in the DEEP model. All abatement is achieved through substitution. The emissions for all gases are provided by the GTAP/EPA databases (Lee, 2002, 2003).

References

Agarwal, A., Narain, S., 1991. Global Warming in an Unequal World: A Case of Environmental Colonialism. Center for Science and Environ- ment, New Delhi. Aldy, J.E., Ashton, J., Baron, R., Bodansky, D., Charnovitz, S., Diringer, E., Heller, T.C., Pershing, J., Shukla, P.R., Tubiana, L., Tudela, F., Wang., X., 2003. Beyond Kyoto: Advancing the international effort against climate change. Pew Center on Global Climate Change, December. Arrhenius, S., 1896. On the influence of carbonic acid in the air upon the temperature of the ground. Philosophical Journal 41, 237–276. Berk, M.M., den Elzen, M.G.J., 2001. Options for Differentiation of Future Commitments in Climate Policy: How to Realise Timely Participation to Meet Stringent Climate Goals? Climate Policy 1,

465–480.

de Campos, C.P., de Muylaert, M.S., Rosa, L.P., 2005. Historical CO 2 emissions and concentrations due to land use change of croplands and pastures by country. Science of Total Environment 346 (1-3),

149–155.

den Elzen, M.G.J., Schaeffer, M., 2002. Responsibility for past and future global warming: uncertainties in attributing anthropogenic climate

change. Climatic Change 54, 29–73.

den Elzen, M.G.J., Fuglestvedt, J., Ho¨hne, N., Trudinger, C., Lowe, J.,

Matthews, B., Romstad, B., de Campos, C., Andronova, N., 2005a.

Analysing countries’ contribution to climate change: Scientific

uncertainties and methodological choices. Environmental Science & Policy 8 (6), 614–636.

den Elzen, M.G.J., Schaeffer, M., Lucas, P.L., 2005b. Differentiating

future commitments on the basis of countries’ relative historical responsibility for climate change: uncertainties in the ‘Brazilian

Proposal’ in the context of a policy implementation. Climatic Change

71 (3), 277–301. Dimaranan, B.V. and McDougall, R.A., 2002. Global Trade, Assistance, and Production: The GTAP 5 Data Base. Center for Global Trade

Analysis, Purdue University.

Fuglestvedt, J.S., Berntsen, T., 1999. A simple model for scenario studies of changes in global climate: version 1.0. Working Paper 1999:02,

CICERO, Oslo, Norway. Fuglestvedt, J.S., Berntsen, T., Godal, O., Skodvin, T., 2000. Climate implications of GWP-based reductions in greenhouse gas emissions. Geophysical Research Letters 27 (3), 409–412. Ho¨hne, N., Blok, K., 2005. Calculating historical contributions to climate change discussing the Brazilian Proposal. Climatic Change 71 (1),

141–173.

Houghton, R.A., 1999. The annual net flux of carbon to the atmos- phere from changes in land use 1850–1990. Tellus 51B,

298–313.

IPCC, 1990. Scientific Assessment of Climate change – Report of Working Group I. Cambridge University Press, Cambridge. IPCC, 1996. Revised 1996 IPCC Guidelines for National Greenhouse Gas Inventories FCCC/SUBSTA/1996/18/Add. 1, 16–18 December. IPCC, 2001. Special Report on Emissions Scenarios. Cambridge University Press, Cambridge. Joos, F., Bruno, M., Fink, R., Stocker, T.F., Siegenthaler, U., Le Que´ re´ , C., Sarmiento, J.L., 1996. An efficient and accurate representation of complex oceanic and biospheric models of anthropogenic carbon uptake. Tellus 48B, 397–417. Kallbekken, S., 2004. A description of the Dynamic analysis of Economic impacts of Environmental Policy (DEEP) model. CICERO Report 2004:01, Oslo. Kallbekken, S., Westskog, H., 2005. Should developing countries take on binding commitments in a climate agreement? An assessment of gains and uncertainty. Energy Journal 26 (3), 41–60. Lee, H.L., 2002. An Emissions data base for integrated assessment of climate change policy using GTAP. GTAP Working Paper (draft), Center for Global Trade Analysis, Purdue University. Lee, H.L., 2003. The GTAP non-CO 2 emissions data base. GTAP Resource #1186, http://www.gtap.agecon.purdue.edu/resources/res_

Marland, G., Boden, T.A., Andres, R.J., 2003. Global, regional, and national CO 2 emissions. Trends: A Compendium of Data on Global Change. Carbon Dioxide Information Analysis Center, Oak Ridge National Laboratory, US Department of Energy, Oak Ridge, TN, USA. Meyer, A., 2000. Contraction and Convergence. The Global Solution to Climate Change. Schumacher briefings 5. Greenbooks for the Schumacher Society, Bristol, USA. Olivier, J.G.J., Berdowski, J.J.M., 2001. Global emissions sources and sinks. In: Berdowski, J., Guicherit, R., Heij, B.J. (Eds.), The Climate System. A.A. Balkema Publishers/Swets & Zeitlinger Publishers, Lisse, The Netherlands, pp. 33–78. Philibert, C., Pershing, J., 2001. Considering the options: climate targets for all countries? Climate Policy 1 (2), 211–227.

ARTICLE IN PRESS

  • 194 N. Rive et al. / Global Environmental Change 16 (2006) 182–194

Philibert, C., Pershing, J., Corfee-Morlot, J., Willems, S., 2003. Evolution of mitigation commitments. OECD and IEA Information Paper,

COM/ENV/EPOC/IEA/SLT(2003)3.

Revelle, R., Suess, H.E., 1957. Carbon dioxide exchange between atmosphere and ocean and the question of an increase of atmospheric CO 2 during the past decades. Tellus 9, 18–27. Rosa, L.P., Ribeiro, S.K., 2001. The Present, past and future contribu- tions to global warming of CO 2 emissions from fuels. Climatic Change 48, 289–308. Rosa, L.P., Ribeiro, S.K., Muylaert, M.S., de Campos, C.P., 2004. Comments on the Brazilian Proposal and contributions to global temperature increase with different climate responses—CO 2 emissions due to fossil fuels, CO 2 emissions due to land use change. Energy Policy 32, 1499–1510. Rose, A., Stevens, B., Edmonds, J., Wise, M., 1998. International equity and differentiations in global warming policy. Environmental and Resource Economics 12, 25–51. Rutherford, T.F., Paltsev, S.V., 2000. GTAP-Energy in GAMS: the dataset and static model. Working Paper No 00-2. Department of Economics, University of Colorado. Schlesinger, M.E., Jiang, K., Charlson, R.J., 1992. Implications of Anthropogenic Atmospheric Sulphate for the Sensitivity of the Climate System, Climate Change and Energy Policy. American Institute of Physics, New York. Schmalensee, R., 1996. Greenhouse policy architecture and institutions. MIT Joint Program on the Science and Policy of Climate Change. Report 13, Cambridge, MA. Shine, K.P., Derwent, R.G., Wuebbles, D.J., Morcrette, J.-J., 1990. Radiative forcing of climate. In: Houghton, J.T., Jenkins, G.J., Ephraums, J.J. (Eds.), Climate Change: The IPCC Scientific Assess- ment. Cambridge University Press, Cambridge, UK, and New York, NY, USA, pp. 41–68.

Siegenthaler, U., Joos, F., 1992. Use of a simple model for studying oceanic tracer distributions and the global carbon cycle. Tellus Series B 44, 186–207. Torvanger, A., Ringius, L., 2002. Criteria for evaluation of burden- sharing rules in international climate policy. International Environ- mental Agreements: Politics, Law and Economics 2 (3), 221–235. Torvanger, A., Twena, M., Vevatne, J., 2004a. Climate policy beyond 2012: a survey of long-term targets and future frameworks. CICERO Report 2004:02, Oslo. Torvanger, A., Berntsen, T., Kallbekken, S., Rive, N., 2004b. A sustainable climate agreement: long-term targets, global agreements, and economic consequences. Presented at Annual EAERE Confer- ence, July 2004, Budapest. Trudinger, C.M., Enting, I.G., 2005. Comparison of formalisms for attributing responsibility for climate change: non-linearities in the Brazilian Proposal. Climatic Change 68, 67–99. UNFCCC, 1992. The United Nations Framework Convention on Climate Change, A/AC.237/18, 9 May. UNFCCC, 1997a. Proposed elements of a protocol to the United Nations framework convention on climate change. In: Proceedings of the Conference in Brazil in Response to the Berlin Mandate. Paper No. 1, Brazil, UNFCCC/AGBM/1997/MISC.1/Add.3 GE.97-, Bonn. UNFCCC, 1997b. Kyoto Protocol to the United Nations Framework Convention on Climate Change, FCCC/CP/1997/7/Add.1, Kyoto. Unruh, G.C., 2000. Understanding carbon lock-in. Energy Policy 28 (12),

817–830.

Van Aardenne, J.A., Dentener, F.J., Olivier, J.G.J., Klein Goldewijk, C.G.M., Lelieveld, J., 2001. A 1 1 degree resolution dataset of historical anthropogenic trace gas emissions for the period 1890-1990. Global Biogeochemical Cycles 15 (4), 909–928. Zhang, Z., 2000. Can China afford to commit itself an emissions cap? An economic and political analysis. Energy Economics 22, 587–614.